Image Signal Processor Remapping for Quantization Error Restoration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image quantization leads to loss of image data due to the conversion of continuous signals into discrete levels, resulting in quantization errors that are not effectively restored in existing image signal processing systems without expanding the number of bits.
Innovation Solution
An image display apparatus and method that includes a signal processor to identify the type of image processing, extract restoration information, and perform remapping to restore quantization errors by extending the number of bits in the image signal processing block without expanding it, using analysis and restoration logic to convert many-to-one mapping to one-to-one mapping and smooth the mapping function.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If image quantization is performed to convert continuous signals into discrete levels, then the image signal can be processed and stored digitally, but quantization errors occur and image data is lost
Solution Approach 1:
The patent applies preliminary action by performing remapping operations before final quantization to reduce quantization errors. The image signal is first remapped using a mapping function that predicts optimal quantization levels, then quantized. This preliminary remapping action prepares the signal in advance to minimize information loss during the necessary quantization process.
Solution Approach 2:
The patent changes the parameter of quantization levels by dynamically determining optimal quantization levels based on image characteristics. Instead of using fixed quantization levels, the system adjusts the number and distribution of quantization levels according to the local variance and statistical properties of the image signal, thereby reducing quantization errors while maintaining digital processing capability.
2Measurement precision
If the number of bits in the image signal is expanded to restore quantization errors, then image quality improves, but hardware and software burdens increase
Solution Approach 1:
The patent changes the parameter of bit depth selectively by maintaining higher bit depth only during intermediate processing and restoration stages while keeping the input and output at standard bit depths. The remapping process uses floating-point or high-precision arithmetic to calculate optimal quantization levels, but the final output remains at the original bit depth, avoiding the need for throughout-system high-bit-depth processing.
Solution Approach 2:
The patent creates a virtual copy of the image signal in the form of a mapping function that stores optimal quantization level information. Instead of actually expanding the bit depth of the image data, the system uses this mapping function copy to guide the quantization process, achieving high-precision restoration effects without the hardware burden of processing actual high-bit-depth image data.
3Loss of information
If remapping is performed using mapping functions to restore quantization errors, then quantization errors are reduced, but processing complexity increases
Solution Approach 1:
The patent segments the image processing into distinct stages: variance calculation, mapping function generation, remapping, and quantization. By dividing the processing into modular segments, each with a specific function, the system manages complexity through functional decomposition. The mapping function generation is performed once per image or image block, and the remapping operation is a simple lookup and adjustment process, making the overall complexity manageable.
Solution Approach 2:
The patent performs preliminary calculation of the mapping function based on image statistics (mean, variance) before the actual quantization process. This preliminary action allows the system to pre-determine optimal quantization levels, so that during the main processing flow, only simple remapping operations are needed rather than complex real-time calculations, reducing processing complexity while maintaining accuracy.
Data Source
AI summary
An object of the present disclosure is to restore a quantization error of an image signal only by expanding the number of bits of a reconstructed image compared to the number of bits of an input image without expanding the number of bits in an image signal processing process. An image display apparatus may include a signal inputter configured to receive an input image signal; an image signal processor configured to perform image processing of the input image signal, to identify a type of image processing by comparing the input image signal and the processed image signal, to obtain restoration information of the processed image signal according to the identified image processing type, and to restore an error of the processed image signal by performing remapping of the processed image signal using the restoration information; and an outputter configured to output the image signal in which the error is restored.